Charged particle beam columns are typically employed in scanning electron microscopy (SEM), which is a known technique widely used in the manufacture of semiconductor wafers, being utilized in a CD metrology tool, the so-called CD-SEM (critical dimension scanning electron microscope), and a defect review SEM (DF-SEM). In SEM, the region of a sample to be examined is two-dimensionally scanned by means of a focused primary beam of electrically charged particles, usually electrons. Irradiation of the sample with the primary electron beam releases secondary (and/or backscattered) electrons. The secondary electrons are released at that side of the sample at which the primary electron beam is incident, and move back to be captured by a detector, which generates an output electric signal proportional to the so-detected electric current. The energy and/or the energy distribution of the secondary electrons is indicative of the nature and composition of the sample.
SEM includes such main constructional parts as an electron beam source (formed with a small tip called “electron gun”), an electron beam column, and a detector unit. The detector unit may be located outside the path of the primary beam propagation through the column, or may be located in the path of the primary beam (the so-called “in-column” or “in-lens” detector). The electron beam column includes, inter alia, a beam focusing/deflecting arrangement formed by a lens assembly and a deflector assembly. The deflection of the primary beam provides for scanning the beam within a scan area on the sample, and also for adjusting incidence of the primary beam onto the sample (an angle of incidence and/or beam shift), as well as directing the secondary beam to the detector.
In SEM, in order to reduce the “spot” size of the electron beam (up to nanometers) and thus increase the image resolution, a highly accelerated electron beam is typically produced using accelerating voltages of several tens of kilovolts and more. Specifically, the electron optic elements are more effective (i.e. produce smaller aberrations) when the electrons are accelerated to high kinetic energy. However, in order to avoid damaging a sample (resist structure and integrated circuit) that might be caused by such a highly energized electron beam, the electron beam is decelerated just prior to impinging onto the sample. Deceleration of the electrons can generally be accomplished by selectively creating a potential difference between the polepiece of a magnetic objective lens and the sample. Alternatively, the same effect can be achieved by actually introducing electrodes having selective potential applied thereto.
Some systems of the kind specified utilize the lens assembly in the form of a combination of a magnetic objective lens and an electrostatic lens, the so-called “compound magnetic-electrostatic lens” (e.g., EP 1238405 and EP 1045425, both assigned to the assignee of the present application). The electrostatic part of the compound magnetic-electrostatic lens is an electrostatic retarding lens (with respect to the primary charged particle beam), and has electrodes held at different potentials, one of the two electrodes being formed by an anode (which is typically in the form of a tube defining a primary beam drift space for the primary beam propagation to the sample) arranged within a magnetic objective lens along its optical axis, and the other electrode being a metallic cup provided below the magnetic objective lens. The sample actually presents the third electrode of the electrostatic lens. The electric field created by the electrostatic lens in the vicinity of the sample appropriately decelerates the primary beam and also facilitates the extraction of secondary charged particles from the sample.
Another known problem of the inspection systems of the kind specified is associated with locating defects (foreign particles) on patterned surfaces. The pattern is typically in the form of a plurality of spaced-apart grooves. To detect the existence of a foreign particle located inside a narrow groove, it is desirable to tilt the scanning beam with respect to the surface, which tilting should be applied to selective locations on the specimen. A tilt mechanism may be achieved by mechanically tilting the sample holder relative to the charged particle beam column, and/or by electronically tilting the primary beam propagation axis. The electronic tilt is implemented by the deflector assembly, which may include one or more deflectors. This may for example be a magnetic deflector integrated into a magnetic objective lens (WO 01/56056), which has an excitation coil and upper and lower polepieces.
U.S. 20040056207, assigned to the assignee of the present application discloses a deflection system including a magnetic deflector (core and polepieces electrically connected to the core) and a polepiece assembly, which has a portion made of a soft magnetic material and is formed with an opening for a charged particle beam passage therethrough. The polepiece assembly is accommodated so as to be at least partly located within the magnetic field of the magnetic deflector to thereby conduct at least a portion of the magnetic field created by the deflector through the polepiece assembly towards the opening. This arrangement increases the effectiveness of deflection and facilitates operation with the tilt mode, by increasing a magnetic field for a given electric current through the excitation coils of the deflector. This allows for obtaining a desirably high deflecting magnetic field within the closest vicinity of the sample at the optical axis of the lens arrangement, without increasing a working distance, also in cases where the compound magnetic-electrostatic lens is used.
Generally, the image resolution of a charged particle beam column can be improved by increasing the anode voltage. However, this might result in a breakdown in the system operation. On the other hand, an increase of a negative voltage (in absolute value) applied to a sample under inspection when operating with a charged particle beam column in which the polepieces of a magnetic deflector are grounded, which is typically the case, is also limited by the breakdown condition at a given working distance. The term “working distance” is typically referred to as a distance between the electrode of the lens arrangement closest to the sample's plane (cap-electrode in the present case) and the sample's plane. This distance should be as small as possible, and the minimal possible working distance is typically defined by an arcing problem.
Yet another characteristic of the charged particle beam column is associated with a need for injection of gas into a vacuum chamber of the column. Generally, a need for the gas supply is associated with the following. Preciseness of measurements, such as CD measurements on semiconductor wafers and especially on lithographic masks (reticles), typically suffers from an effect of charging of the sample's surface by a scanning beam of charged particles, which causes an image drift. The gas supply into the vicinity of the sample and ionization by the scanning beam allow for reducing negative charging of the sample's surface sue to the precipitation of positive ions onto the sample's surface, and thus allow for increasing the precision of CD measurements.
The gas supply unit typically used in a DR-SEM includes a special nozzle tube for feeding the gas. The nozzle is mounted on the side surface of a conical objective lens such that the outlet opening of the nozzle is located proximate of the beam opening of the objective lens. This configuration is, however, unsuitable for a CD-SEM that typically utilizes a substantially flat cap-electrode and requires a smaller working distance (about 0.8 mm, instead of 1.5 mm used in DR-SEM). The conventional configuration of a gas supply unit used in the CD-SEM suffers from that disabling the cap electrode breaks the tightness connection between the electrode and the gas supply unit, because a gas tube thereof is connected to a holder separate from the cap electrode. Moreover, the connection zone is too small and therefore cannot be sufficiently reliable. Also, this configuration does not allow for using the so-called post-tilt deflector aimed at providing an on-axis tilt of a primary beam.
An example of the conventional gas supply system is described in U.S. Pat. No. 6,182,605. An apparatus for particle beam induced modification of a specimen includes a source for generating a particle beam, nozzles for supplying a gas in the region of the specimen, and electrodes which can be supplied with a variable voltage. These electrodes are forming a tube and the nozzles, being integrated in the electrodes, are leading into the tube.
U.S. Pat. No. 6,555,815, assigned to the assignee of the present application, discloses a charged particle beam column, where charging of the specimen is avoided or reduced by injecting inert gas onto the sample's surface. In order to avoid interactions with the electron optics, various embodiments are disclosed for providing a rotationally symmetrical nozzles and/or electrodes. Additionally, embodiments are disclosed wherein a plurality of gas conduits are arranged in a rotationally symmetrical manner. Alternatively, the conduit is incorporated into an element of the electron optics, such as the magnetic lens. Also, in order to reduce or eliminate interaction of the electrons with the gas molecules, embodiments are disclosed wherein the gas is pulsated, rather than continually injected.